1. Field of the Invention
The present invention is related to calixarenes and their preparation. More specifically, it is directed to a new and useful method for the preparation of known as well as of new calixarenes. The invention further aims at the new calixarenes obtained by the process of the invention.
Calixarenes are cyclic phenolic oligomers wherein two adjacent phenol rings are linked together by an ortho-ortho CH.sub.2 bridge or an analogous homo CH.sub.2 bridge. Some calixarenes have already been known, in particular the tetramers, hexamers and octamers as well as several calixarenes having an odd number of rings in their molecule. It seems that calixarenes having less than three phenolic rings are not accessible.
The calixarenes have found several applications in industry; they are for example used as starting materials for solid foams, as capturing agents for other, small molecules, as reactants for confining waste or dangerous solids, etc.
2. Description of the Prior Art
As it has already been mentioned above, some calixarenes are already known, see, e.g., the work of Gutsche et al. in Org. Syn. 1989, 68, 234; Niederl and Vogel in J. Am. Chem. Soc. 1990, 62, 2512, and Hogberg in J. Am. Chem. Soc. 1980, 102, 6046. p-Methylcalix[4]arene has already been synthesized in 1956 by a ten-step synthesis method using brominated intermediates. This methods has been taken over and developed in order to prepare linear phenolic oligomers which are then cyclized for yielding calixarenes having an even or odd number of phenolic rings, sometimes differently substituted, or asymmetric calixarenes (see, e.g., V. Bohmer et al., Makromol. Chem. 180, 2503-06 (1979), who however only discloses cyclic tetramers). Linear phenolic oligomers are at least two phenolic rings linked together in the ortho positions by a methylene group. In all cases, the syntheses are long-lasting and give very low yields only.
In 1981, a critical review of the two most common methods for synthesizing calixarenes has been published by C. D. Gutsche and co-workers in J. Am. Chem. Soc. 1981, 103, 3782-92. These two methods are a one-step method, called Munch method, and a three-step method, called Zinke method.
In the Munch method, a mixture containing p-tert-butylphenol, paraformaldehyde, and potassium hydroxide in an approximate ratio of 45:75:1 is heated during 4 hours in a xylene solution that is approximately 1.3 M in phenol, using a Dean and Stark trap to remove water. Typically, a colorless, xylene insoluble product is obtained in about 75% yield. This product contains linear phenolic oligomers and p-tert-butylcalix[8]arene (45%) whereas the 6- and 4-calixarene homologues can be isolated in small amounts (i.e. 9.5% and 9%, respectively) from the mother liquor.
The Zinke method comprises warming a mixture of the above mentioned starting materials during about 45 hours to 50-55.degree. C. and then for about 2 hours to 110-120.degree. C. Then, the reaction mixture is acidified, the separated solids are removed, and the washed and dried solids are heated for about 2 hours in a high-boiling liquid to about 210-220.degree. C. This method principally yields cyclic tetramer and small amounts of hexamer and octamer. It has also been proposed by Gutsche, see the document cited in the preceding paragraph, to combine the first two steps of this method.
U.S. Pat. No. 4,098,717 discloses pre-formed phenolic compounds, substituted in one ortho position or in one ortho and in one meta position by a methylol group, as starting substances for a cyclization. The present invention does not use such starting compounds but para substituted phenols. Moreover, the reference clearly describes that a cyclization of these preformed phenolic formaldehyde derivatives under neutral conditions does not yield calixarenes but cyclic ethers, and that under mildly basic conditions, cyclic resins are formed, thus no calixarenes.
European patent application No. 0,447,977 discloses a method for the preparation of cyclic phenolic tetramers by heating a precondensate of p-alkylphenols and formaldehyde or an alkylcalix[8]arene together with an aqueous alkali metal hydroxide during about 1 to 10 hours at 217 to 257.degree. C. in a hydrocarbon solvent. This variant of the Zinke method yields cyclic tetramer only in a yield of 50 to 60% only, and furthermore, the said precondensates are isolated prior to the final condensation.
Another one-step method for the synthesis of calixarenes and bishomooxacalixarenes, using a rather strongly basic medium, is disclosed in Chemical Abstracts 97, 1982, No. 197914h, without indicating reaction parameters such as temperature and duration. Furthermore, such a one-step method, using a basic medium too, is disclosed in Chemical Abstracts 92, 1980, No. 22301a where the presence of bishomooxacalix[4]arenes in the reaction product is mentioned.
Finally, attempts have been made to synthesize calixarenes comprising an odd number of rings in their molecule. Thus, A. Ninagawa suggests in Makromol. Chem., Rapid Commun. 3, 1982, 65-67 the isolation of the cyclic pentamer from the reaction product of the first two steps of a modified Zinke method. According to Y. Nakamoto [Makromol. Chem., Rapid Commun. 3, 705-07 (1982)], the heptamer was obtained in low yield from the solid reaction product of p-tert-butylphenol, paraformaldehyde and potassium hydroxide after refluxing this mixture for 30 hours in dioxane.
The methods of the prior art discussed above suffer from several severe drawbacks. No one of these methods can be used to prepare a desired calixarene in high yields on an industrial scale since on one hand, these known methods are either too time consuming and require a great number of separating, washing, drying and isolating operations, or they only give calixarenes in an industrially unsatisfactorily low yield, on the other hand, or both.